KITE: A Knowledge-Injected Temporal Encoder for Time-Series Event Forecasting

Key points

• Forecasting rare, high-cost events in multivariate time-series systems is hard because their precursors are seldom observed, so a purely data-driven model has little to learn from.1
• The engineers who run these systems, however, hold knowledge that precedes such events, information absent from the scarce labels.1
• KITE is a switchable, attributable route to contextual conditioning, and a substrate for scaling toward cross-domain, pretrained time-series foundation world models that carry their governing structure explicitly.1

Sources (1)

• [1]KITE: A Knowledge-Injected Temporal Encoder for Time-Series Event Forecasting
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 8, 09:35 PM

  We call this a semantic prior and inject it into a self-supervised event forecaster through two switchable mechanisms, a covariate mechanism and an attention-bias mechanism, each the identity at its null, so one architecture serves every prior type with no per-dataset architecture search.

  Forecasting rare, high-cost events in multivariate time-series systems is hard because their precursors are seldom observed, so a purely data-driven model has little to learn from.

Extractive summary: sentences quoted from the sources.